Re: Opinions sought: AudioUnit Rendering
Re: Opinions sought: AudioUnit Rendering
- Subject: Re: Opinions sought: AudioUnit Rendering
- From: "Douglas Wright" <email@hidden>
- Date: Wed, 6 Mar 2002 14:32:10 -0800
- Organization: Bias, Inc.
Are there not AudioUnits with different size buffer requirements? How do
you handle the case when the request is for more data than the size of the
buffer of zeros?
Returning a NULL buffer is sure to crash any units that don't check to see
if there is actually a buffer there.
If the AU is required to return as many samples as was asked for, then do.
But you can return a bytecount that represents the non-silent data in the
buffer rather than just mimicking the requested size. Old units that
expected to get what they asked for will work and use up cpu processing
silence, and new AUs can do the right thing.
The buffer of zeros does save some time setting buffers full of zeroes, but
it requires more compares in every audio unit to check if it's the BOZ and
that the size of BOZ is big enough to respond to the request.
dug
----- Original Message -----
From: "Doug Wyatt" <email@hidden>
To: "CoreAudio API" <email@hidden>
Sent: Wednesday, March 06, 2002 9:14 AM
Subject: Re: Opinions sought: AudioUnit Rendering
>
While this sort of protocol has its uses, in this case it would be a
>
seriously non-backwardly-compatible change to the existing AudioUnit
>
API. It would require that all existing AudioUnits check to see whether
>
they actually got all the data they asked for, though the existing API
>
contract for RenderSlice is that the AU must provide as much data as was
>
asked for. So we can't do that.
_______________________________________________
coreaudio-api mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/coreaudio-api
Do not post admin requests to the list. They will be ignored.